ALEXANDER AGASSIZ 441 
rain water and atmospheric erosion leached out depressions in its cen- 
tral parts and finally the sea entered, forming a lagoon surrounded by 
a ring of detached islets of elevated limestone. In other cases the 
crater rims have washed away and a ring of coral reefs now marks the 
site of the old voleanic ridge. 
According to Agassiz the coral reefs of to-day, in the Fiji Islands, 
form only a crust of moderate thickness upon a base of old limestone 
or volcanic rock. The present corals form only fringing reefs along 
the shores, and the contours of the atolls and barrier reefs are thus due 
to causes which acted at the time when the islands were elevated late in 
the Tertiary period. 
In so great an archipelago as that of the Fijis with more than 270 
islands there must be many details of reef formation the elucidation of 
which requires more prolonged study than Agassiz was enabled to 
devote to them in his visit of less than three months; for example, he 
was puzzled to explain the great thickness—1,000 feet and more—of 
the elevated limestones; for reef-corals do not grow at depths greater 
than about 120 feet. Could these enormous accumulations have been 
formed by coral reefs during a period of slow subsidence, as Darwin 
had assumed, or were they merely the talus of broken fragments which 
had rolled down the sea-ward sides from the outer edges of the reef, 
or were they formed by a slow accumulation of limestone fragments 
and shells of marine creatures other than corals which had lived upon 
the bottom more than a thousand feet beneath the surface and gradu- 
- ally built up a vast mass of limestone, as was the case with the great 
submerged Pourtalés Plateau off the Florida coast? He had in mind 
the fact that even in the richest coral reef regions the masses of broken 
shells and fragments of calcareous plants are commonly vastly greater 
than the bulk of the corals, for the corals grow only here and there 
over the limestone flats, and flourish luxuriantly only on their sea- 
ward slopes. 
Were such a reef to form during a long period of slow subsidence, 
and then become elevated above the sea we should find only an occasional 
coral here and there imbedded in a great mass of hmestone. This is 
the appearance presented by some of these elevated limestone cliffs of 
the Pacific islands, while others appear to be walls of non-coral-bearing 
limestone capped above with a crust of corals. In many cases, however, 
the corals they once contained have disappeared and been replaced by 
calcite or dolomite. These elevated limestones soon become very hard 
when exposed to the atmosphere, for they become coated by a dense 
veneer which rings with a clinker-like sound when struck with a ham- 
mer. One sometimes finds shells of the giant clam, 7'ridacna, imbedded 
in this hard limestone and elevated above the sea, and yet the nacre 
of the shell is still white and polished, thus proving that the rock 
